CPS 296.1
Normal-form games
Vincent Conitzer
conitzer@cs.duke.edu
Rock-paper-scissors
Column player (also
known as player 2)
(simultaneously)
chooses a column
0, 0 -1, 1 1, -1
Row player
(also known as
player 1)
chooses a row
A row or column is
called an action or
(pure) strategy
1, -1 0, 0 -1, 1
-1, 1 1, -1 0, 0
Row player’s utility is always listed first, column player’s second
Zero-sum game: the utilities in each entry sum to 0 (or a constant)
Three-player game would be a 3D table with 3 utilities per entry, etc.
Matching pennies (~penalty kick)
L
R
L 1, -1
-1, 1
R -1, 1
1, -1
“Chicken”
• Two players drive cars towards each other
• If one player goes straight, that player wins
• If both go straight, they both die
S
D
D
S
D
D
S
S
0, 0 -1, 1
1, -1 -5, -5
not zero-sum
Rock-paper-scissors – Seinfeld variant
MICKEY: All right, rock beats paper!
(Mickey smacks Kramer's hand for losing)
KRAMER: I thought paper covered rock.
MICKEY: Nah, rock flies right through paper.
KRAMER: What beats rock?
MICKEY: (looks at hand) Nothing beats rock.
0, 0 1, -1 1, -1
-1, 1 0, 0 -1, 1
-1, 1 1, -1 0, 0
Dominance
• Player i’s strategy si strictly dominates si’ if
– for any s-i, ui(si , s-i) > ui(si’, s-i)
• si weakly dominates si’ if
– for any s-i, ui(si , s-i) ≥ ui(si’, s-i); and
– for some s-i, ui(si , s-i) > ui(si’, s-i)
strict dominance
weak dominance
-i = “the player(s)
other than i”
0, 0 1, -1 1, -1
-1, 1 0, 0 -1, 1
-1, 1 1, -1 0, 0
Prisoner’s Dilemma
• Pair of criminals has been caught
• District attorney has evidence to convict them of a
minor crime (1 year in jail); knows that they
committed a major crime together (3 years in jail)
but cannot prove it
• Offers them a deal:
– If both confess to the major crime, they each get a 1 year reduction
– If only one confesses, that one gets 3 years reduction
confess
confess
don’t confess
don’t confess
-2, -2 0, -3
-3, 0 -1, -1
“Should I buy an SUV?”
accident cost
purchasing cost
cost: 5
cost: 3
cost: 5
cost: 5
cost: 8
cost: 2
cost: 5
cost: 5
-10, -10
-7, -11
-11, -7
-8, -8
Mixed strategies
• Mixed strategy for player i = probability
distribution over player i’s (pure) strategies
• E.g.,1/3
, 1/3
, 1/3
• Example of dominance by a mixed strategy:
1/2
1/2
3, 0 0, 0
0, 0 3, 0
1, 0 1, 0
Usage:
σi denotes a
mixed strategy,
si denotes a pure
strategy
Checking for dominance by mixed strategies
• Linear program for checking whether strategy si* is
strictly dominated by a mixed strategy:
• maximize ε
• such that:
– for any s-i, Σsi psi ui(si, s-i) ≥ ui(si*, s-i) + ε
– Σsi psi = 1
• Linear program for checking whether strategy si* is
weakly dominated by a mixed strategy:
• maximize Σs-i[(Σsi psi ui(si, s-i)) - ui(si*, s-i)]
• such that:
– for any s-i, Σsi psi ui(si, s-i) ≥ ui(si*, s-i)
– Σsi psi = 1
Iterated dominance
• Iterated dominance: remove (strictly/weakly)
dominated strategy, repeat
• Iterated strict dominance on Seinfeld’s RPS:
0, 0 1, -1 1, -1
-1, 1 0, 0 -1, 1
-1, 1 1, -1 0, 0
0, 0 1, -1
-1, 1 0, 0
Iterated dominance: path (in)dependence
Iterated weak dominance is path-dependent:
sequence of eliminations may determine which
solution we get (if any)
(whether or not dominance by mixed strategies allowed)
0, 1
1, 0
0, 0
0, 0
1, 0
0, 1
0, 1
1, 0
0, 0
0, 0
1, 0
0, 1
0, 1
1, 0
0, 0
0, 0
1, 0
0, 1
Iterated strict dominance is path-independent: elimination
process will always terminate at the same point
(whether or not dominance by mixed strategies allowed)
Two computational questions for
iterated dominance
• 1. Can a given strategy be eliminated using iterated
dominance?
• 2. Is there some path of elimination by iterated
dominance such that only one strategy per player
remains?
• For strict dominance (with or without dominance by
mixed strategies), both can be solved in polynomial
time due to path-independence:
– Check if any strategy is dominated, remove it, repeat
• For weak dominance, both questions are NP-hard
(even when all utilities are 0 or 1), with or without
dominance by mixed strategies [Conitzer, Sandholm 05]
– Weaker version proved by [Gilboa, Kalai, Zemel 93]
Two-player zero-sum games revisited
• Recall: in a zero-sum game, payoffs in each entry sum to zero
– … or to a constant: recall that we can subtract a constant from
anyone’s utility function without affecting their behavior
• What the one player gains, the other player loses
0, 0 -1, 1 1, -1
1, -1 0, 0 -1, 1
-1, 1 1, -1 0, 0
Note: a general-sum k-player
game can be modeled as a zerosum (k+1)-player game by adding
a dummy player absorbing the
remaining utility, so zero-sum
games with 3 or more players
have to deal with the difficulties of
general-sum games; this is why
we focus on 2-player zero-sum
games here.
Best-response strategies
• Suppose you know your opponent’s mixed strategy
– E.g., your opponent plays rock 50% of the time and scissors
50%
•
•
•
•
•
What is the best strategy for you to play?
Rock gives .5*0 + .5*1 = .5
Paper gives .5*1 + .5*(-1) = 0
Scissors gives .5*(-1) + .5*0 = -.5
So the best response to this opponent strategy is to
(always) play rock
• There is always some pure strategy that is a best
response
– Suppose you have a mixed strategy that is a best response;
then every one of the pure strategies that that mixed strategy
places positive probability on must also be a best response
How to play matching pennies
Them
L
Us
•
•
•
•
•
R
L 1, -1
-1, 1
R -1, 1
1, -1
Assume opponent knows our mixed strategy
If we play L 60%, R 40%...
… opponent will play R…
… we get .6*(-1) + .4*(1) = -.2
What’s optimal for us? What about rock-paper-scissors?
Matching pennies with a sensitive target
Them
L
Us
R
L 1, -1
-1, 1
R -2, 2
1, -1
• If we play 50% L, 50% R, opponent will attack L
– We get .5*(1) + .5*(-2) = -.5
• What if we play 55% L, 45% R?
• Opponent has choice between
– L: gives them .55*(-1) + .45*(2) = .35
– R: gives them .55*(1) + .45*(-1) = .1
• We get -.35 > -.5
Matching pennies with a sensitive target
Them
L
Us
R
L 1, -1
-1, 1
R -2, 2
1, -1
• What if we play 60% L, 40% R?
• Opponent has choice between
– L: gives them .6*(-1) + .4*(2) = .2
– R: gives them .6*(1) + .4*(-1) = .2
• We get -.2 either way
• This is the maximin strategy
– Maximizes our minimum utility
Let’s change roles
Them
L
Us
R
L 1, -1
-1, 1
R -2, 2
1, -1
• Suppose we know their strategy
• If they play 50% L, 50% R,
– We play L, we get .5*(1)+.5*(-1) = 0
• If they play 40% L, 60% R,
– If we play L, we get .4*(1)+.6*(-1) = -.2
– If we play R, we get .4*(-2)+.6*(1) = -.2
• This is the minimax strategy
von Neumann’s minimax
theorem [1927]: maximin
value = minimax value
(~LP duality)
Minimax theorem [von Neumann 1927]
• Maximin utility: maxσi mins-i ui(σi, s-i)
(= - minσi maxs-i u-i(σi, s-i))
• Minimax utility: minσ-i maxsi ui(si, σ-i)
(= - maxσ-i minsi u-i(si, σ-i))
• Minimax theorem:
maxσi mins-i ui(σi, s-i) = minσ-i maxsi ui(si, σ-i)
• Minimax theorem does not hold with pure
strategies only (example?)
Practice games
20, -20 0, 0
0, 0 10, -10
20, -20
0, 0
10, -10
0, 0
10, -10
8, -8
Solving for minimax strategies
using linear programming
• maximize ui
• subject to
for any s-i, Σsi psi ui(si, s-i) ≥ ui
Σsi psi = 1
Can also convert linear programs to two-player
zero-sum games, so they are equivalent
General-sum games
• You could still play a maximin strategy in generalsum games
– I.e., pretend that the opponent is only trying to hurt you
• But this is not rational:
0, 0
1, 0
3, 1
2, 1
• If Column was trying to hurt Row, Column would play Left, so
Row should play Down
• In reality, Column will play Right (strictly dominant), so Row
should play Up
• Is there a better generalization of maximin strategies in zerosum games to general-sum games?
Nash equilibrium
[Nash 50]
• A vector of strategies (one for each player) is called
a strategy profile
• A strategy profile (σ1, σ2 , …, σn) is a Nash
equilibrium if each σi is a best response to σ-i
– That is, for any i, for any σi’, ui(σi, σ-i) ≥ ui(σi’, σ-i)
• Note that this does not say anything about multiple
agents changing their strategies at the same time
• In any (finite) game, at least one Nash equilibrium
(possibly using mixed strategies) exists [Nash 50]
• (Note - singular: equilibrium, plural: equilibria)
Nash equilibria of “chicken”
D
S
S
D
D
S
D
S
0, 0 -1, 1
1, -1 -5, -5
• (D, S) and (S, D) are Nash equilibria
– They are pure-strategy Nash equilibria: nobody randomizes
– They are also strict Nash equilibria: changing your strategy will make
you strictly worse off
• No other pure-strategy Nash equilibria
Nash equilibria of “chicken”…
D
S
D
S
0, 0 -1, 1
1, -1 -5, -5
• Is there a Nash equilibrium that uses mixed strategies? Say, where player 1
uses a mixed strategy?
• Recall: if a mixed strategy is a best response, then all of the pure strategies
that it randomizes over must also be best responses
• So we need to make player 1 indifferent between D and S
• Player 1’s utility for playing D = -pcS
• Player 1’s utility for playing S = pcD - 5pcS = 1 - 6pcS
• So we need -pcS = 1 - 6pcS which means pcS = 1/5
• Then, player 2 needs to be indifferent as well
• Mixed-strategy Nash equilibrium: ((4/5 D, 1/5 S), (4/5 D, 1/5 S))
– People may die! Expected utility -1/5 for each player
The presentation game
Presenter
Audience
Pay
attention (A)
Do not pay
attention (NA)
Put effort into
presentation (E)
Do not put effort into
presentation (NE)
4, 4
0, -2
-16, -14
0, 0
• Pure-strategy Nash equilibria: (A, E), (NA, NE)
• Mixed-strategy Nash equilibrium:
((1/10 A, 9/10 NA), (4/5 E, 1/5 NE))
– Utility 0 for audience, -14/10 for presenter
– Can see that some equilibria are strictly better for both players than other
equilibria, i.e. some equilibria Pareto-dominate other equilibria
The “equilibrium selection problem”
• You are about to play a game that you have never
played before with a person that you have never met
• According to which equilibrium should you play?
• Possible answers:
– Equilibrium that maximizes the sum of utilities (social
welfare)
– Or, at least not a Pareto-dominated equilibrium
– So-called focal equilibria
• “Meet in Paris” game - you and a friend were supposed to meet in
Paris at noon on Sunday, but you forgot to discuss where and you
cannot communicate. All you care about is meeting your friend.
Where will you go?
– Equilibrium that is the convergence point of some learning
process
– An equilibrium that is easy to compute
–…
• Equilibrium selection is a difficult problem
Some properties of Nash equilibria
• If you can eliminate a strategy using strict
dominance or even iterated strict dominance, it
will not occur (i.e., it will be played with
probability 0) in every Nash equilibrium
– Weakly dominated strategies may still be played in
some Nash equilibrium
• In 2-player zero-sum games, a profile is a Nash
equilibrium if and only if both players play
minimax strategies
– Hence, in such games, if (σ1, σ2) and (σ1’, σ2’) are
Nash equilibria, then so are (σ1, σ2’) and (σ1’, σ2)
• No equilibrium selection problem here!
How hard is it to compute one
(any) Nash equilibrium?
• Complexity was open for a long time
– [Papadimitriou STOC01]: “together with factoring […] the
most important concrete open question on the boundary
of P today”
• Recent sequence of papers shows that computing
one (any) Nash equilibrium is PPAD-complete (even
in 2-player games) [Daskalakis, Goldberg, Papadimitriou 2006; Chen,
Deng 2006]
• All known algorithms require exponential time (in the
worst case)
What if we want to compute a Nash
equilibrium with a specific property?
• For example:
– An equilibrium that is not Pareto-dominated
– An equilibrium that maximizes the expected social welfare (= the
sum of the agents’ utilities)
– An equilibrium that maximizes the expected utility of a given player
– An equilibrium that maximizes the expected utility of the worst-off
player
– An equilibrium in which a given pure strategy is played with positive
probability
– An equilibrium in which a given pure strategy is played with zero
probability
– …
• All of these are NP-hard (and the optimization questions are
inapproximable assuming P ≠ NP), even in 2-player games
[Gilboa, Zemel 89; Conitzer & Sandholm IJCAI-03/GEB-08]
Search-based approaches (for 2 players)
• Suppose we know the support Xi of each
player i’s mixed strategy in equilibrium
– That is, which pure strategies receive positive
probability
• Then, we have a linear feasibility problem:
– for both i, for any si  Si - Xi, pi(si) = 0
– for both i, for any si  Xi, Σp-i(s-i)ui(si, s-i) = ui
– for both i, for any si  Si - Xi, Σp-i(s-i)ui(si, s-i) ≤ ui
• Thus, we can search over possible supports
– This is the basic idea underlying methods in
[Dickhaut & Kaplan 91; Porter, Nudelman, Shoham AAAI04/GEB08]
• Dominated strategies can be eliminated
Solving for a Nash equilibrium
using MIP (2 players)
[Sandholm, Gilpin, Conitzer AAAI05]
• maximize whatever you like (e.g., social welfare)
• subject to
– for both i, for any si, Σs-i ps-i ui(si, s-i) = usi
– for both i, for any si, ui ≥ usi
– for both i, for any si, psi ≤ bsi
– for both i, for any si, ui - usi ≤ M(1- bsi)
– for both i, Σsi psi = 1
• bsi is a binary variable indicating whether si is
in the support, M is a large number
Lemke-Howson algorithm (1-slide sketch!)
RED
BLUE
player 2’s utility as
function of 1’s mixed
strategy
RED
•
•
•
BLUE
GREEN
ORANGE
1, 0
0, 2
0, 1
1, 0
player 1’s utility as
function of 2’s mixed
strategy
redraw both
GREEN
best-response strategies
ORANGE
unplayed strategies
Strategy profile = pair of points
Profile is an equilibrium iff every pure strategy is either a best response or
unplayed
I.e. equilibrium = pair of points that includes all the colors
– … except, pair of bottom points doesn’t count (the “artificial equilibrium”)
•
Walk in some direction from the artificial equilibrium; at each step, throw out the
color used twice
Correlated equilibrium [Aumann 74]
• Suppose there is a trustworthy mediator who has offered to
help out the players in the game
• The mediator chooses a profile of pure strategies, perhaps
randomly, then tells each player what her strategy is in the
profile (but not what the other players’ strategies are)
• A correlated equilibrium is a distribution over pure-strategy
profiles so that every player wants to follow the
recommendation of the mediator (if she assumes that the
others do so as well)
• Every Nash equilibrium is also a correlated equilibrium
– Corresponds to mediator choosing players’ recommendations
independently
• … but not vice versa
• (Note: there are more general definitions of correlated
equilibrium, but it can be shown that they do not allow you to
do anything more than this definition.)
A correlated equilibrium for “chicken”
D
S
0,
0
-1,
1
D 20%
40%
S
1, -1
-5, -5
40%
0%
• Why is this a correlated equilibrium?
• Suppose the mediator tells the row player to Dodge
• From Row’s perspective, the conditional probability that Column was told
to Dodge is 20% / (20% + 40%) = 1/3
• So the expected utility of Dodging is (2/3)*(-1) = -2/3
• But the expected utility of Straight is (1/3)*1 + (2/3)*(-5) = -3
• So Row wants to follow the recommendation
• If Row is told to go Straight, he knows that Column was told to Dodge, so
again Row wants to follow the recommendation
• Similar for Column
A nonzero-sum variant of rock-paperscissors (Shapley’s game [Shapley 64])
•
•
•
•
0, 0
0, 1
1, 0
0
1/6
1/6
1, 0
0, 0
0, 1
1/6
0
1/6
0, 1
1, 0
0, 0
1/6
1/6
0
If both choose the same pure strategy, both lose
These probabilities give a correlated equilibrium:
E.g. suppose Row is told to play Rock
Row knows Column is playing either paper or scissors (50-50)
– Playing Rock will give ½; playing Paper will give 0; playing Scissors will give ½
• So Rock is optimal (not uniquely)
Solving for a correlated equilibrium
using linear programming (n players!)
• Variables are now ps where s is a profile of pure
strategies
• maximize whatever you like (e.g., social welfare)
• subject to
– for any i, si, si’, Σs-i p(si, s-i) ui(si, s-i) ≥ Σs-i p(si, s-i) ui(si’, s-i)
– Σs ps = 1